Date of Award

Spring 2023

Project Type

Thesis

Program or Major

Mechanical Engineering

Degree Name

Master of Science

First Advisor

May-Win Thein

Second Advisor

May-Win Thein

Third Advisor

Jennifer Jacobs

Abstract

NASA studies indicate that 68% of the Earth’s fresh water exists in the form of snow andice. As such, analyzing global snow fall patterns is a useful tool with which scientists can extract the quantity of fresh water present in both the atmosphere and on the ground at any given time. The goal of this research is to leverage autonomous Unpiloted Aerial Vehicles (UAVs) to measure snow depth on the forest floor via sub-canopy flight. To enable such remote sensing missions, overhead Light Detection And Ranging (LiDAR) scans are used to aid in pre-determined UAV flight path planning. This results in autonomous sub-canopy missions that are able to avoid obstacles (e.g., trees, branches, and flora) and provide optimal LiDAR-based snow depth measurements. The A-star (A*) algorithm is the chosen path planning method for this research and is used to determine appropriate flight plans for multi-UAV missions.Ox Bow Farm, Kingman Farm, and Thompson Farm are evaluated and have sub-canopy tree density between 60-90%, 20-42% and 25-55% respectively. Proof-of-concept testing is performed at the University of New Hampshire Kingman Farm in Madbury, New Hampshire. Field tests show that this method is viable for under-canopy snow depth measurements when tree density is below 20%. In addition to the added efficiency of an autonomous multi-UAV mission (as opposed to a single, remotely operated UAVs), the resulting sub-canopy photogrammetry results, from which snow depth measurements can be extracted, are shown to provide improved ability to capture snow as compared to above-canopy flights.

Download

Share

COinS